Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I downloaded all 1.6M posts on Bluesky

worthdoingbadly.com

Post reply on HN
Print view View original

151–160 of 294 posts

Re: I downloaded all 1.6M posts on Bluesky

#151

Post by mkl »

ChrisMarshallNY wrote: post #119
btown wrote: post #46

Given that Bluesky makes it this easy to download data, it's quite alarming that the graph of who blocks & mutes who is fully public and easy to extract into a database: https://atproto.com/lexicons/app-bsky-graph#appbskygraphgetb... On Twitter, blocking a toxic user does not notify them - while they can query the block status of one profile at a time, they can never get a full list of people who block them. But it w…

I'm writing a fairly basic social media app, right now. The target demographic is very security/privacy-sensitive, so we have discussions about stuff like this, all the time. If you block someone, they have no idea, except that you disappear from their radar. Even that, only came about, after a long discussion. Before, blocking someone simply meant that we could not see them, but they could see us. My location is som…

> Even when it is visible, it is "fuzzed," so that the location anyone (other than you) can see, is a random one, within a 10Km square.

You may want to expand that in lower population density areas. If there's only one little town in that 10km square the fuzzing does almost nothing.

Re: I downloaded all 1.6M posts on Bluesky

#152

Post by astrange »

brvsft wrote: post #89
dragonwriter wrote: post #72

Earlier quoted context omitted.

You can just not click on Bluesky links, no one is forcing you to read articles you aren’t interested in.

[flagged]

This didn't happen, but more importantly Twitter was far more proactive about resisting law enforcement requests for private info than other social media companies, which is the actual thing to be worried about.

Current Twitter ownership is allowing India's government to censor things worldwide and is obviously compromised in China as well.

Re: I downloaded all 1.6M posts on Bluesky

#153

Post by Zetice »

eli wrote: post #130
btown wrote: post #123

Earlier quoted context omitted.

The problem is when B comes across a A post and reposts it to B’s followers with a call to action to harass A. If you don’t give A the ability to avoid this by influencing A’s presence on B’s feed, A will quit the platform entirely, and B will have been rewarded (with engagement) and incentivized to stir up mobs in the future. It creates a toxic environment. Blocking is an effective speed bump to this happening!

Strong user safety features are at odds with a federated protocol. IMHO federation is a mistake.

It's so much easier to build this into the app (which is what you're saying) rather than try to solve this at the proto level, I'm honestly surprised they even tried to include it into the protocol.

Then again, I shouldn't be surprised. There's a really large trans community active on Bluesky right now, and they've been extremely vocal about what's important to them, with blocking being the thing demanded the loudest (butts being #2).

The dev team blew up their schedule to deliver blocking because of this.

Re: I downloaded all 1.6M posts on Bluesky

#154

Post by ChrisMarshallNY »

mkl wrote: post #151
ChrisMarshallNY wrote: post #119

Earlier quoted context omitted.

I'm writing a fairly basic social media app, right now. The target demographic is very security/privacy-sensitive, so we have discussions about stuff like this, all the time. If you block someone, they have no idea, except that you disappear from their radar. Even that, only came about, after a long discussion. Before, blocking someone simply meant that we could not see them, but they could see us. My location is som…

> Even when it is visible, it is "fuzzed," so that the location anyone (other than you) can see, is a random one, within a 10Km square. You may want to expand that in lower population density areas. If there's only one little town in that 10km square the fuzzing does almost nothing.

That’s a good suggestion. I’d probably calculate the spread, based on the density of the community members, as opposed to outside stats.

It’s not urgent, at present, but it could be done. See my answer to another point.

Re: I downloaded all 1.6M posts on Bluesky

#156

Post by sodality2 »

capableweb wrote: post #137
TaylorAlexander wrote: post #133

Earlier quoted context omitted.

On twitter if a profile is locked (cannot be viewed publicly and followers must be approved) it’s very difficult for a blocked user to ever see a post. Even for a public profile if someone is blocked they can’t use the normal QT option and they can’t reply, which creates much more friction for follow on abuse.

> On twitter if a profile is locked (cannot be viewed publicly and followers must be approved) it’s very difficult for a blocked user to ever see a post. It's really not. Incognito window usually work, or something like nitter. Not to mention all the various archives of public tweets. > Even for a public profile if someone is blocked they can’t use the normal QT option and they can’t reply, which creates much more fr…

> Incognito window usually work, or something like nitter

They won't if the profile is locked. It means you have to be logged in to check you aren't on their blocklist (or maybe even it's equivalent to account being private - I'm not up to date on Twitter).

Re: I downloaded all 1.6M posts on Bluesky

#157

Post by TaylorAlexander »

capableweb wrote: post #137
TaylorAlexander wrote: post #133

Earlier quoted context omitted.

On twitter if a profile is locked (cannot be viewed publicly and followers must be approved) it’s very difficult for a blocked user to ever see a post. Even for a public profile if someone is blocked they can’t use the normal QT option and they can’t reply, which creates much more friction for follow on abuse.

> On twitter if a profile is locked (cannot be viewed publicly and followers must be approved) it’s very difficult for a blocked user to ever see a post. It's really not. Incognito window usually work, or something like nitter. Not to mention all the various archives of public tweets. > Even for a public profile if someone is blocked they can’t use the normal QT option and they can’t reply, which creates much more fr…

> Incognito window usually work, or something like nitter.

This does not work for locked profiles, which cannot be viewed unless logged in.

Re: I downloaded all 1.6M posts on Bluesky

#158

Post by chillbill »

btown wrote: post #46

Given that Bluesky makes it this easy to download data, it's quite alarming that the graph of who blocks & mutes who is fully public and easy to extract into a database: https://atproto.com/lexicons/app-bsky-graph#appbskygraphgetb... On Twitter, blocking a toxic user does not notify them - while they can query the block status of one profile at a time, they can never get a full list of people who block them. But it w…

You’re being too dramatic about it… it’s not that big of a deal.

Re: I downloaded all 1.6M posts on Bluesky

#159

Post by AkshatJ27 »

crossroadsguy wrote: post #138
capableweb wrote: post #97

Earlier quoted context omitted.

Yes! And no, last one :(

That is cool. People get only one invite? Anyway I have added a temp email in my profile. Just in case one lands in my inbox.

Everyone gets an invite every 2 weeks. The staff can give anyone more invites as per their will

Re: I downloaded all 1.6M posts on Bluesky

#160

Post by notatoad »

VancouverMan wrote: post #80
btown wrote: post #46

Given that Bluesky makes it this easy to download data, it's quite alarming that the graph of who blocks & mutes who is fully public and easy to extract into a database: https://atproto.com/lexicons/app-bsky-graph#appbskygraphgetb... On Twitter, blocking a toxic user does not notify them - while they can query the block status of one profile at a time, they can never get a full list of people who block them. But it w…

Maybe it'd be better to just not have any blocking/muting built into such a system at all. If User A doesn't want to see posts from User B, that's fine. User A can have his client filter them out locally, prior to when they'd otherwise be displayed. Nobody else has to know this is happening. I don't think that User A should be permitted to prevent User B from replying to User A's posts, which in turn prevents User C…

there's no reason you can't filter privately in the client as well, if you have a client that wants to prioritize privacy. but blocking/muting seems like a reasonable thing to include in the protocol, for clients that are prioritizing resources and not ingesting more posts than they need to and as a signal for algorithms to determine users who are frequently blocked or muted.

for somebody like myself, i'm just some dude nobody cares about. my twitter blocklist was major public figures who i didn't want to see retweets of in my feed. donald trump wasn't going to get angry at me because i blocked him, it's fine if that's included in some public list.

Post reply on HN